Openbrowser Ai

OpenBrowser is a Python framework for automating browser tasks from plain text instructions. A developer describes a task in natural language, and an agent powered by a large language model drives a real browser to carry it out, such as logging in to a site and checking notifications. The documentation shows a small set of building blocks. An Agent takes a task, a browser and a model; a Browser object controls which Chrome instance is used; and chat classes such as ChatGoogle wrap the supported language model providers. The guides explain how to test locally with an existing Chrome profile and its saved logins, how to keep a dedicated persistent user data directory between runs, and how to run headless for production. A browser can also be a remote instance reached over the Chrome DevTools Protocol, which suits server deployments. Best practice notes cover keeping API keys in environment variables. The project is aimed at developers who want to script web workflows, scrape or operate sites behind a login, or embed a browsing agent into their own applications. A quick start and a list of supported model providers are provided, along with a documentation index that can be fetched in machine readable form. The documentation lists no paid plans or prices. The library is used with the user's own model provider account, so costs depend on the model chosen.